Amplitude CEO Spence Skates appeared on the Y Combinator podcast to share a vivid account of the organizational restructuring and mindset shifts that a public company underwent to adapt to the AI era. He candidly reveals the bold decisions he made to abandon the SaaS development routines that had driven past success and transform into an AI-native company, along with the internal conflicts that arose. He also shares hard-won lessons and insights gained in the process of growing from a hands-on "founder mode" into an executive capable of efficiently running a large organization.

1. The Journey from AI Skeptic to AI Evangelist

Early in the video, Spence confesses that he and the Amplitude team were quite skeptical of AI just one to two years ago. In 2022 and early 2023, the AI boom was underway but its substance was unclear.

An anecdote from a board meeting at the time is quite entertaining -- investors and finance staff were asking "AI is hot, so what's our strategy?" while the people actually doing the work were frustrated.

"One of our executives actually came to me directly and asked, 'Spence, what's our AI strategy?' I told him that's not the right way to think about what we do. I was almost thinking, 'If you think you know what it is, why don't you go do it yourself?'"

His co-founder Jeffrey, in particular, was deeply frustrated by the "grifting" prevalent in the AI industry. There were many cases of flashy talk with no useful results to show. But the tide turned starting in late 2024. They witnessed the enormous productivity gains AI brought to software engineering (especially through tools like Cursor).

"AI model capabilities are very jagged. Some things they do amazingly well, others they're terrible at. (...) But seeing the transformative effect AI had on software engineering changed our thinking. It became clear that developers using these tools were far more productive."

From October 2024, Amplitude began its transformation in earnest -- hiring legendary Silicon Valley engineering leader Wade Chambers and acquiring YC-backed startup Command AI to accelerate the pace of change.

2. You Must Abandon Old Success Formulas to Become AI-Native

Spence explains the critical difference between the "traditional SaaS development approach" and the "AI product development approach." The formula that led Amplitude to success over the past decade was an iterative routine of asking customers what they wanted and building it for them. But he emphasizes that this approach no longer works in the AI era.

"SaaS is the greatest business model and product delivery system in history. You go to customers and ask what they want, prioritize, build, deploy, and repeat. But because AI capabilities are still incomplete, if you go to customers and ask 'What do you need?' they don't even know what's possible."

Much like Henry Ford's famous observation that "if I had asked people what they wanted, they would have said a faster horse," customers struggle to imagine what AI can do. Therefore, the approach must now be technology-first: understanding what the technology can do before building the product.

To this end, Amplitude created "AI Week" for company-wide training.

"We used the phrase 'burning the boats.' There's no going back. (...) The first step wasn't just building AI products -- it was getting existing teams to use AI tools and trust their capabilities."

Through this process, they had to part ways with some leaders who could not adapt to AI or clung to old methods. Instead, AI-passionate internal engineers began generating innovative ideas (such as AI Visibility and Ask AI) in a bottom-up fashion.

3. Even Large Companies Cannot Escape the "Death of SaaS"

On the hot industry debate of "Will AI kill SaaS?", Spence acknowledges the changes AI will bring while pointing out that "reliability" remains crucial in the B2B space.

"Business workflows require a high level of performance guarantees. You can't put data into a CRM and have it say 'There's an 80% chance it was saved.' (...) That's why the key challenge is building products well where users can edit and review, rather than fully automated agents."

At the same time, he warns that simple feature-level services will be rapidly displaced by AI. For example, the "AI Visibility" feature that Amplitude recently released for free illustrates how capabilities that once commanded premium prices are becoming commoditized thanks to AI.

He also amusingly points out how vulnerable large companies like Google are in the B2B market, and how great the opportunity is for startups.

"This might sound like a joke, but every area Google is trying to compete in is an opportunity. Google is historically the worst B2B company ever. (...) Google Docs and Workspace -- innovation is too slow and conservative. When you look at what companies like Notion are doing, it's really exciting."

4. Founder Mode vs. Executive Mode

One of the core themes of this video is the transformation from "Founder" to "Executive." Spence explains how as the company grew, he had to move away from the approach of personally solving everything.

"As a founder, your job is always to rush to the hardest problem in the business and lead from the front. (...) But when you become an executive at a large company, you can't lead from the front everywhere. You have to be much more disciplined about time management, and you have to say 'no' to most things."

The most painful realization in this process was that he had to become the very type of "big company executive" he used to despise.

"You end up realizing you've become the person you hated the most. I used to mock big-company executives who didn't do the work themselves and just evaluated other people's work. There were reasons for that all along. You have to accept that role."

He also came to deeply appreciate why organizational hierarchy is necessary. Not everything can be done horizontally -- leaders with responsibility and authority are needed.

"Honestly, being a big-company executive is much easier. You have more resources, you can create leverage, and you actually do less work. (...) But you need completely different tools and skills from your founder days. That transition was the hardest part for me."

5. Conclusion: "The Power to Persevere" and the Future of Analytics

Finally, Spence emphasizes the importance of "intrinsic motivation" for those who dream of starting a company or are going through tough times. Chasing money or fame alone will not sustain you to the end.

"From a rational standpoint, quitting at the one- or two-year mark might be the right call. But the people who succeeded didn't quit for one reason or another. (...) Clarifying why you do this work -- that top node -- is the most important thing."

Amplitude now aims to go beyond being a simple analytics tool and completely reinvent the analytics paradigm through AI. He is confident that the "Cursor moment for analytics" will arrive within the next two years.

"In the coming years, a reinvention of analytics will happen. We want to be the protagonist leading that change."

In Closing

This video goes beyond the story of simply adopting AI technology to address the inevitable changes and pain that leaders of growing companies must face.

• For developers: Pay attention to how AI is changing coding and product development practices.
• For founders: Be clear about your answer to "Why am I doing this?" and accept that your role must evolve as the company grows.
• For executives: Consider whether past success formulas are becoming shackles in the new technology era.